Click here to return to the Basic DSP page
There are two kinds of subtraction supported.
The Division block allows you to divide two incoming signals. The division is performed using the Newton-Raphson iteration.
For a sample design using this block, see the Basic DSP example.
To use this block:
The Newton-Raphson iteration is performed according to the equation:
This block lets you select the precision of the algorithm, whether to compute 3 or 4 iterations. There's a tradeoff between number of instructions and accuracy of computation for divisors (values of pin 2) less than 0.1. This means simply that fewer iterations are not as precise as more, but the more iterations the more instructions are entailed, with less room for programming the current DSP.
Below are error graphs for the 3- (below top) and 4-iteration (below bottom) algorithms, showing a difference of approximately two orders of magnitude:
This module can be used to divide a complex signal with another complex signal.
(x+iy) / (u+iv) = (xu + yv) + i(yu - xv) / (u^2 + v^2)
Note:- The Context Menu “Input Source” option added to this module to select the algorithm for Complex FFT and Real FFT as shown below
Name | Format [int/dec] - [control/audio] | Function Description |
---|---|---|
Pin 0: Operand 1 | complex | Input complex signal 1 |
Pin 1: Operand 2 | complex | Input Complex signal 2 |
Name | Format [int/dec] - [control/audio] | Function Description |
---|---|---|
Pin 0: Output Signal | Complex | Complex output signal |
Input pins growth not supported.
ADAU145x (Block Schematic only)
This can be used to divide complex signals